# 🔎 Product Research Brief

> Duplicate this page for every product you look at. Fill it top to bottom. If you get to the verdict and it is a NO, throw the page away and start another one. That is the job — most products are a no.

**Product:** 
**Date started:** 
**Status:** ☐ Researching ☐ Validated ☐ Running ☐ Killed ☐ Scaling

---

## 1. What is it

| Field | Your answer |
| --- | --- |
| Product in one line |  |
| Type | ☐ Handmade ☐ Wow factor ☐ Problem solving |
| Who actually buys it |  |
| Who *pays* for it (often not the same person) |  |
| Rough selling price |  |
| Can I source it? | ☐ AliExpress ☐ Private supplier ☐ Not checked yet |
| Supplier cost |  |

**Why I think this can work (2 sentences, no hype):**



---

## 2. The accounts running it

Pull these off your burner feed. One row per account. You need at least 2 before you judge anything.

| Account | Best post (plays) | Followers | Posting how long | Product page | Notes |
| --- | --- | --- | --- | --- | --- |
| @ |  |  |  |  |  |
| @ |  |  |  |  |  |
| @ |  |  |  |  |  |

**Score the product on the ceiling, not the floor.** Rank on the single best post and how many videos cleared 1M. Ignore follower count and account average — neither predicts the next hit.

- Best post across all of them: 
- How many of their videos cleared 1M: 
- Newest account started posting: 

---

## 3. The per-mille test

Test-order their cheapest item. The confirmation email gives you an order number.

| Field | Value |
| --- | --- |
| Their main product price | $ |
| Order number I received |  |
| Order numbers start at (Shopify default 1000) | 1000 |
| **Orders so far** (current − start) |  |
| **Revenue** (orders × price) | $ |
| Their total views |  |
| **Per mille** (revenue ÷ views in millions) | $ |

**Read it:**
- **$5,000+ per million** → strong, it converts
- **$2,000–5,000** → decent, worth it with a fresh angle
- **Under $2,000** → weak, views but no sales

**Verdict:** 

**If you can't test-order,** check the comments instead. Are people asking *where can I buy this*, or just saying *this is cool*? Only the first one is demand.

- Demand comments seen: ☐ Yes ☐ No — examples: 

---

## 4. Can I actually win it?

Saturation is **not** a competitor count. It is whether a proven product-and-concept combo is already going viral repeatedly. You can still win a taken product if **any one** of these is true:

- [ ] They are not that viral yet (best post under ~1M)
- [ ] There is a concept they have not run
- [ ] There is a platform they skipped (TikTok / IG / Facebook)
- [ ] I can run a variant (cat mug → dog mug)
- [ ] I can bring a character none of them are using

**Which one is my opening:**



**What they have already run** (so I don't repeat it):

| Concept they used | How it did | Would I be copying? |
| --- | --- | --- |
|  |  |  |
|  |  |  |

---

## 5. THE VERDICT

Score each 0–2. Be harsh.

| Check | Score |
| --- | --- |
| Real demand (buy comments or a strong per-mille) | /2 |
| Can actually go viral (stops the scroll) | /2 |
| Buyers who have money |  /2 |
| I have a clear opening (section 4) | /2 |
| I can source it at a workable margin | /2 |
| **TOTAL** | **/10** |

**8+ → run it. 6–7 → only if the opening is genuinely strong. Under 6 → drop it and find another.**

**Decision:** ☐ RUN IT ☐ PASS

If passing, one line on why, so you don't reconsider the same product in three weeks:



---

## 6. Concepts to run

Lift these off winners. Don't invent from scratch — copy what already keeps people rewatching, then change one thing.

| # | Concept / hook line | Where I saw it | Plays | Structure | My change |
| --- | --- | --- | --- | --- | --- |
| 1 |  |  |  |  |  |
| 2 |  |  |  |  |  |
| 3 |  |  |  |  |  |
| 4 |  |  |  |  |  |
| 5 |  |  |  |  |  |

**Structure shorthand:** handmade = hook → build → result · wow = hook → buildup → reveal · problem = hook → problem → before/after

**Shape rules that came out of 160 winners:**
- 8–16 seconds total, target 12
- 4–7 beats, roughly 2 seconds each
- First beat is the hook, 3.0–3.5s
- **One build beat, or none.** Two or three is the worst structure there is.
- Every video carries an on-screen hook line, burned in at edit time
- Write for the laugh. Humour outperforms shock and transformation.

---

## 7. My character

The character attracts the audience. Match it to whoever buys, and to a look **no competitor running this product is using** — that freshness is what makes the algorithm treat your video as new.

**Character DNA — write it once, never reword it:**

```
[age range] [gender] [ethnicity], [hair — be very specific about style and how it sits],
[skin tone + a real texture detail like visible pores], [face shape],
[one signature accessory]
```

**Mine:**

```

```

| Field | Answer |
| --- | --- |
| Why this character for this buyer |  |
| What look competitors are using (so I avoid it) |  |
| One character or rotate? | ☐ One (handmade / personal brand) ☐ Rotate (wow / problem) |
| Master character image saved at |  |

⚠️ Never describe the character as attractive, beautiful or model-perfect. Real and slightly imperfect is what converts.

---

## 8. Accounts + posting

| Platform | Handle | Made? | Primed? | First post |
| --- | --- | --- | --- | --- |
| TikTok | @ | ☐ | ☐ |  |
| Instagram | @ | ☐ | ☐ |  |
| Facebook | @ | ☐ | ☐ |  |

**Prime a new account before posting:** search your niche, like and save 10–20 videos, wait about an hour.

**Posting rules:** 3–5 videos a day, 2–3 hours apart. Get 5–10 up before you drop the product link. Never delete a video.

---

## 9. Tracking

| # | Date | Platform | Concept | Views (24h) | Views (72h) | Comments | Notes |
| --- | --- | --- | --- | --- | --- | --- | --- |
| 1 |  |  |  |  |  |  |  |
| 2 |  |  |  |  |  |  |  |
| 3 |  |  |  |  |  |  |  |
| 4 |  |  |  |  |  |  |  |
| 5 |  |  |  |  |  |  |  |
| 6 |  |  |  |  |  |  |  |
| 7 |  |  |  |  |  |  |  |
| 8 |  |  |  |  |  |  |  |
| 9 |  |  |  |  |  |  |  |
| 10 |  |  |  |  |  |  |  |

---

## 10. Kill or scale

Check this after 5–10 solid videos. A product gets about 3 days.

- [ ] **SCALE** — a video passed 5–10k views, still climbing, and I'm getting where-can-I-buy comments → **build the store now**, then split-test the video that hit
- [ ] **KEEP TESTING** — under 5 videos posted, or there's movement but no clear signal
- [ ] **NEW ACCOUNT** — views stuck at 200–500 and the content is genuinely good → it's targeting, make a fresh account
- [ ] **KILL** — 5–10 solid videos, no traction, no buy comments → drop it, find the next one

**Decision + date:** 

**What I learned (fill this in even when you kill it):**



---

## 📋 The research prompt

Paste this into your Claude with the accounts you found. This is the step that does the heavy lifting.

```
I'm researching a product for AI organic dropshipping.

Product: [WHAT IT IS]
Type: [handmade / wow factor / problem solving]
Accounts running it: [PASTE LINKS]
Their best videos: [PASTE LINKS]

Work through this and fill in my research brief:

1. The product, and who actually buys it vs who pays for it.
2. For each account: best post, followers, how long they've been posting,
   and how many videos cleared 1M. Score on the ceiling, not the average.
3. Estimate the per-mille at their price if you can, and tell me what
   the comments say about demand — where-can-I-buy vs this-looks-cool.
4. Can I win it? Only yes if they aren't that viral yet, there's an
   untried concept, a platform they skipped, or I can run a variant.
   Name my specific opening.
5. Five concepts I could run, with the hook line, the beat-by-beat
   structure, and the one thing I'd change on each.
6. The best character to attract this buyer, and a look no competitor
   running this product is using.

Score it out of 10 and tell me straight whether to run it or pass.
Don't pad it. If it's a pass, say so.
```

---

## 🗂 How to use this in Notion

1. **Notion → Import → Markdown** and select this file, or just paste the whole thing into a blank page.
2. Turn this page into a **template button** or a database template so every new product starts from a clean copy.
3. Keep the killed ones. A folder of ten passes is how you get fast at spotting the eleventh.